Cartridge spindle for a slow speed grinder / diamond lap

[jeremysf]

Ok, after a bit of a break in the action, I’m working on the spindle again.

I chucked the spindle up in my 6” 3-jaw chuck with copper soft jaws and dialed it in as true as I am capable of (around a few tenths).

I faced the small “nub” on the spindle nose that I was unable to reach due to turning between centers, but was careful not to touch the bulk of the face to try to preserve as much of the squareness of that face to the axis of rotation as possible.

I then drilled out the spindle nose to 3/8” tapping size (drill Q) to about 1” deep, and then undersize drilled and reamed a 1/2” shoulder.

It’s always a mistake for me to try to ream precise 1/2” holes on the lathe. For some reason I always come out 0.001” over. I don’t have this issue with other reamer sizes, and I have played with feeds and speeds in the past, and I’m starting to suspect my 1/2” import reamer may not be the best.

I cleaned up the bottom of the shoulder with a small inset boring bar and then tapped the hole with a 3/8-24 tap. And with that, the spindle body is complete!

Next up I turned the bolt that seats in the shoulder and holds the 6” aluminum backing plate and diamond wheel to the spindle nose. I bought the aluminum backing plate and a selection of diamond wheels off of ebay for cheap. I could have machined the aluminum backing plate, but I was able to buy it machined for cheaper than I can buy the raw stock!

I started by chucking up some 1” mystery mild steel. It doesn’t machine very well, but I don’t have any 12L14 that diameter.

I turned the OD down to 0.800” and then turned down a 0.350” shoulder to a 1/2” OD. Next, I turned down about 3/4” to an OD of 3/8” and machined a 0.051” deep groove immediately behind the shoulder using my 1/8” parting tool for thread relief.

I then single point threaded the 3/8-24 threads and parted the bolt off, leaving a 0.210” thick chunk of the 0.800” OD at the end.

Over to the mill, I put the bolt in a 5C collet block to machine two wrench flats onto the 0.800” head of the bolt. I picked a 5/8” spacing on the flats, and that was the bolt done!

Time for a test assembly of the spindle nose, aluminum backing plate, diamond plate and the retaining bolt.

Looking good!

Next up is the spindle housing and the front retaining nut that holds the spindle in the housing.

The retaining nut in particular is a bit intimidating as it’s a very spindly little part with some precise external threading required. The book suggests machining a custom split collet, but I’m going to see if I can get away with machining the OD and ID in the 4-jaw, parting it off as a last step.

Progress!

 

[7milesup]

Looking good!!

 

[jeremysf]

Latest progress: machined the spindle housing.

First up, cut a 4.5” chunk of 2.125” round (12L14 free machining steel) on the horizontal bandsaw:

Chucked it up in the 4-jaw and indicated it to run true, both at the chuck side, and then tapping it in with a dead blow hammer; rinse and repeat until it was dead true. I would estimate probably less than 0.0002” TIR, which I’m guessing is about the limit of my lathe and chuck.

I faced the end, center drilled, added tail support with a live center and turned down the part of the OD sticking out:

Next I removed the tail support and step drilled up to the largest drill I had at the time which was 0.5” as deep as I could (about 2.25” deep). I have since bought a 1” drill and some morse taper adaptor sleeves for my MT4 tailstock.

With so much stick out and not really enough space for a steady or follow rest, I opted for very light cuts with the boring bar (0.030”) to open the 0.5” hole up to 1.25” which took…some patience and time. The dimension is not a critical one as it’s just a clearance dimension, but I figured it would be good practice.

I periodically checked my progress, mostly to practice hitting dimension. I used the “balanced cut” technique for this, and before checking with an inside mic, always a spring pass to compensate for the deflection of the smallish boring bar I was using. I do have larger boring bars, but didn’t bother with them due to the light cuts and large stick out.

I also periodically checked the runout of the part with a dial indicator, but only once had to re-true it.

I then expanded the bore for the back bearing recess. The recess is supposed to support a light push fit of the back bearing. I used the back bearing to check the fit, but was able to hit my dimension sharp using the “balanced cuts” method rather than trying to sneak up on it. The light push fit is to accommodate (theoretical) movement of the back bearing due to thermal expansion. I think that is unnecessary given the low RPM of my application.

I put the part on the surface plate, blued up the last 1” of the unmachined OD and scribed a line with the height gage at 4.125”.

I put the part back in the 4-jaw flipped end for end to work on the unmachined end. I took my time to realllly true up the part, only indicating off of the machined OD at the most extreme ends of that machined surface as was exposed.

I faced the part for length, again with light cuts due to the large stick out. Center drilled, tail support, and turned the unmachined OD to match the other machined OD.

I removed the tail support and drilled a 0.5” hole to connect through to the 1.25” bore, and then opened up the hole with the boring bar to 1.25”, now all the way through the part.

Next I opened up another clearance recess to 1.5” about 2” deep.

From here on out the pucker factor was extremely high.

I opened up the bore to be a stiff push fit for the OD of the two front bearings to a depth of the two bearings plus 0.125” (for the front bearing retaining nut to thread into).

I am still learning how to judge extremely close fits. I hit my dimension sharp using balanced cuts, but it’s so hard to tell how close the fit is for real. I used a bearing to test, but was deathly afraid of getting the bearing stuck in the bore, with good reason, I did get it stuck a few times. In the end I just crossed my fingers that I was close enough without being too much of an interference fit.

I relieved the base is the inside shoulder by a few thou so that the inner race of the front bearing would seat properly against the inside shoulder.

Last operation was the threads for the front bearing retaining nut. Using the threading tool, I machined a 0.025” groove, 0.054” depth of cut (double thread depth for 24tpi) about 0.125” deep into the bore as both a place for the threads to land, and so that the inner race of the front most bearing would not be sitting on threads.

I then cut 0.125” of 24tpi threads (about three full threads) with super duper light passes, like 0.005” or less with multiple periodic spring passes. I cut the threads under power, but using jog function of my VFD. I was super impressed by how clean the threads came out. I think I’m finally getting the hang of single pointing internal threads. I left the half nut engaged the whole time and backed out my cross slide and used reverse jog for each pass. I used the threading tool to put a light chamfer on the ID is the end of the part.

Lastly, some emory paper on the thread crests, and about 10 minutes of a 240->320->400 grit emory progression on the OD of the spindle housing and that’s the part done!

Last part to machine is the front bearing retaining nut that clamps the outer races of the front bearings against the shoulder of the spindle housing.

Getting close!

 

[jeremysf]

I finally wrapped up my spindle project!

The last major piece remaining to create was the front bearing retaining nut.

The major challenge of this piece was that it was quite thin.

I chucked up a piece of 0.325” thick, 2.375” 12L14 round in the 4-jaw chuck. I gripped about 0.125” of the piece in the jaws, and dialed it in as concentric as I could, given that the round was unmachined.

I then faced it with light cuts (
Needless to say, I went slow, triple checked for interference between the tools and chuck as this was very close work to the chuck (within 0.010”).

I next turned down 0.157” to match the ID of the spindle housing to prepare for cutting the threads. Before cutting the threads, I used a 0.032” grooving tool (i.e. a thin HSS parting blade) to cut the thread relief up against the shoulder.

I ground a custom HSS external threading tool on my Deckel grinder to be able to thread within < 0.032” of the shoulder. I kept the half nut engaged the whole time and cut the threads using the jog function of the VFD.

I checked the fit of the external threads against the internal threads of the spindle body without removing the nut in progress from the chuck. Sadly, for some reason I just checked the start of the threads, which would come back to bite me later.

Next, I bored out the inside of the nut, chamfered and deburred and removed from the chuck.

With all lathe operations of parts complete, I attempted a test assembly and disaster! The nut would not thread into the body far enough to clamp the front bearings.

I spent a bunch of time trying to clean up potential binding points with needle files and emory paper, bluing up the parts to check for how and where it was binding, etc to no avail.

Finally, I hit the bullet and chucked the nut back up in the 4-jaw, indicating on the ID and squaring it up using a dial test indicator on the inner face.

I picked up the threads using the compound and cross slide and 10x magnification, and then deepened the threads by 0.002” and…SUCCESS.

I then milled six castellations in the nut using my rebuilt import dividing head and a 0.125” end mill.

The result came out very nice, and I was able to use a spanner to tighten down the nut, resulting in zero axial play in the spindle.

Next up, I broached the keyway in the pulley, and milled the matching keyway on the spindle shaft.

My last step was to mill wrench flats in the pulley nut, and…done!

I measured the TIR to < 0.001”. Pretty good for my first attempt at a spindle!

I’m now working on the spindle and motor mount for the grinder. I’ll update the thread with the finished tool.

Thanks for following along!

 

[extropic]

Nice Spindle. The thread describes your process very clearly, which should be helpful to many.

 

[jeremysf]

Final steps for my slow speed grinder / lap.

I created a pair of spindle clamps / mounts from aluminum. I first cleaned up the two aluminum blanks in the mill and used a spotting drill to mark out the location of the 2.125” bore that the spindle housing will mount in. I used a cheap 2” shell mill off Amazon with aluminum specific inserts and wow, surface finish!

I then chucked up the parts in the 4-jaw on the lathe and step drilled the hole, with a final drill size of 1” using my new large morse taper drills and reducer sleeves to fit it in my MT4 tail stock. I then bored the holes to final dimension of 2.125” using a boring bar.

Next, I sawed the parts, bisecting the hole, and cleaned up the sawn faces on the mill, removing enough material to leave a gap between them to aid in the clamping action.

I then drilled and tapped the taller bottom parts for some long 10-32 socket head cap screws, and clearance drilled matching holes in the top parts, finishing the work off with countersinks using an end mill.

Next, I drilled and tapped a series of holes along the bottom edges to be able to secure to it the base plate, a piece of 1/8” steel.

I then laid out and drilled holes in the base plate for the spindle clamps, the motor, the bridge rectifier and a set of rubber feet. The base plate was some scrap mystery steel which I cleaned up with a file and some scotch brite.

The motor is an eBay find: a geared 200rpm Carter shunt motor that takes 115v DC. I picked a shunt motor as it is has constant speed under load. It did mean I needed to order a $15 bridge rectifier to convert 115v AC to DC.

Lastly, I machined a block of aluminum as the work rest, square on the bottom and with a 5 degree angle on the top.

And with that, the project is more or less complete! I’m going to do a bit of cable management, likely bolting on a small L bracket to hold the power cable strain relief and power switch, but I put it into operation this afternoon and it works great!

Since the spindle, bearings and motor are more or less sealed, and the diamond lap plate in operation gets lubricated with WD-40 virtually eliminating any carbide dust, I’m not going fabricate a case.

Plus, wouldn’t want to hide all the machining work that went into it!